fix: ensure proper usage of DataDir (#12300)

- GetObject() should always use a common dataDir to
  read from when it starts reading, this allows the
  code in erasure decoding to have sane expectations.

- Healing should always heal on the common dataDir, this
  allows the code in dangling object detection to purge
  dangling content.

These both situations can happen under certain types of
retries during PUT when server is restarting etc, some
namespace entries might be left over.

cmd/api-errors.go

@@ -766,7 +766,7 @@ var errorCodes = errorCodeMap{
 	},
 	ErrSlowDown: {
 		Code:           "SlowDown",
-		Description:    "Please reduce your request",
+		Description:    "Resource requested is unreadable, please reduce your request rate",
 		HTTPStatusCode: http.StatusServiceUnavailable,
 	},
 	ErrInvalidPrefixMarker: {

cmd/bitrot-streaming.go

@@ -146,6 +146,11 @@ func (b *streamingBitrotReader) ReadAt(buf []byte, offset int64) (int, error) {
 		streamOffset := (offset/b.shardSize)*int64(b.h.Size()) + offset
 		if len(b.data) == 0 && b.tillOffset != streamOffset {
 			b.rc, err = b.disk.ReadFileStream(context.TODO(), b.volume, b.filePath, streamOffset, b.tillOffset-streamOffset)
+			if err != nil {
+				logger.LogIf(GlobalContext,
+					fmt.Errorf("Error(%w) reading erasure shards at (%s: %s/%s), will attempt to reconstruct if we have quorum",
+						err, b.disk, b.volume, b.filePath))
+			}
 		} else {
 			b.rc = io.NewSectionReader(bytes.NewReader(b.data), streamOffset, b.tillOffset-streamOffset)
 		}
@@ -153,7 +158,6 @@ func (b *streamingBitrotReader) ReadAt(buf []byte, offset int64) (int, error) {
 			return 0, err
 		}
 	}
-
 	if offset != b.currOffset {
 		// Can never happen unless there are programmer bugs
 		return 0, errUnexpected

cmd/erasure-decode.go

@@ -32,7 +32,6 @@ type parallelReader struct {
 	readers       []io.ReaderAt
 	orgReaders    []io.ReaderAt
 	dataBlocks    int
-	errs          []error
 	offset        int64
 	shardSize     int64
 	shardFileSize int64
@@ -49,7 +48,6 @@ func newParallelReader(readers []io.ReaderAt, e Erasure, offset, totalLength int
 	return &parallelReader{
 		readers:       readers,
 		orgReaders:    readers,
-		errs:          make([]error, len(readers)),
 		dataBlocks:    e.dataBlocks,
 		offset:        (offset / e.blockSize) * e.ShardSize(),
 		shardSize:     e.ShardSize(),
@@ -173,7 +171,6 @@ func (p *parallelReader) Read(dst [][]byte) ([][]byte, error) {
 				// This will be communicated upstream.
 				p.orgReaders[bufIdx] = nil
 				p.readers[i] = nil
-				p.errs[i] = err
 
 				// Since ReadAt returned error, trigger another read.
 				readTriggerCh <- true
@@ -198,19 +195,8 @@ func (p *parallelReader) Read(dst [][]byte) ([][]byte, error) {
 		return newBuf, nil
 	}
 
-	if countErrs(p.errs, nil) == len(p.errs) {
-		// We have success from all drives this can mean that
-		// all local drives succeeded, but all remote drives
-		// failed to read since p.readers[i] was already nil
-		// for such remote servers - this condition was missed
-		// we would return instead `nil, nil` from this
-		// function - it is safer to simply return Quorum error
-		// when all errs are nil but erasure coding cannot decode
-		// the content.
-		return nil, errErasureReadQuorum
-	}
-
-	return nil, reduceReadQuorumErrs(context.Background(), p.errs, objectOpIgnoredErrs, p.dataBlocks)
+	// If we cannot decode, just return read quorum error.
+	return nil, errErasureReadQuorum
 }
 
 // Decode reads from readers, reconstructs data if needed and writes the data to the writer.

cmd/erasure-healing-common.go

@@ -200,7 +200,6 @@ func disksWithAllParts(ctx context.Context, onlineDisks []StorageAPI, partsMetad
 	object string, scanMode madmin.HealScanMode) ([]StorageAPI, []error) {
 	availableDisks := make([]StorageAPI, len(onlineDisks))
 	dataErrs := make([]error, len(onlineDisks))
-
 	inconsistent := 0
 	for i, meta := range partsMetadata {
 		if !meta.IsValid() {
@@ -270,6 +269,9 @@ func disksWithAllParts(ctx context.Context, onlineDisks []StorageAPI, partsMetad
 			if dataErrs[i] == nil {
 				// All parts verified, mark it as all data available.
 				availableDisks[i] = onlineDisk
+			} else {
+				// upon errors just make that disk's fileinfo invalid
+				partsMetadata[i] = FileInfo{}
 			}
 			continue
 		}
@@ -291,6 +293,9 @@ func disksWithAllParts(ctx context.Context, onlineDisks []StorageAPI, partsMetad
 		if dataErrs[i] == nil {
 			// All parts verified, mark it as all data available.
 			availableDisks[i] = onlineDisk
+		} else {
+			// upon errors just make that disk's fileinfo invalid
+			partsMetadata[i] = FileInfo{}
 		}
 	}
 

cmd/erasure-healing.go

@@ -28,7 +28,6 @@ import (
 
 	"github.com/minio/madmin-go"
 	"github.com/minio/minio/cmd/logger"
-	"github.com/minio/minio/pkg/bucket/lifecycle"
 	"github.com/minio/minio/pkg/sync/errgroup"
 )
 
@@ -267,10 +266,27 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 
 	// List of disks having latest version of the object er.meta
 	// (by modtime).
-	latestDisks, modTime, dataDir := listOnlineDisks(storageDisks, partsMetadata, errs)
+	_, modTime, dataDir := listOnlineDisks(storageDisks, partsMetadata, errs)
+
+	// make sure all parts metadata dataDir is same as returned by listOnlineDisks()
+	// the reason is its possible that some of the disks might have stale data, for those
+	// we simply override them with maximally occurring 'dataDir' - this ensures that
+	// disksWithAllParts() verifies same dataDir across all drives.
+	for i := range partsMetadata {
+		partsMetadata[i].DataDir = dataDir
+	}
+
+	// List of disks having all parts as per latest metadata.
+	// NOTE: do not pass in latestDisks to diskWithAllParts since
+	// the diskWithAllParts needs to reach the drive to ensure
+	// validity of the metadata content, we should make sure that
+	// we pass in disks as is for it to be verified. Once verified
+	// the disksWithAllParts() returns the actual disks that can be
+	// used here for reconstruction. This is done to ensure that
+	// we do not skip drives that have inconsistent metadata to be
+	// skipped from purging when they are stale.
+	availableDisks, dataErrs := disksWithAllParts(ctx, storageDisks, partsMetadata, errs, bucket, object, scanMode)
 
-	// List of disks having all parts as per latest er.meta.
-	availableDisks, dataErrs := disksWithAllParts(ctx, latestDisks, partsMetadata, errs, bucket, object, scanMode)
 	// Loop to find number of disks with valid data, per-drive
 	// data state and a list of outdated disks on which data needs
 	// to be healed.
@@ -336,6 +352,7 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 		// File is fully gone, fileInfo is empty.
 		return defaultHealResult(FileInfo{}, storageDisks, storageEndpoints, errs, bucket, object, versionID, er.defaultParityCount), err
 	}
+
 	// If less than read quorum number of disks have all the parts
 	// of the data, we can't reconstruct the erasure-coded data.
 	if numAvailableDisks < result.DataBlocks {
@@ -405,7 +422,7 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 		result.ParityBlocks = latestMeta.Erasure.ParityBlocks
 
 		// Reorder so that we have data disks first and parity disks next.
-		latestDisks = shuffleDisks(availableDisks, latestMeta.Erasure.Distribution)
+		latestDisks := shuffleDisks(availableDisks, latestMeta.Erasure.Distribution)
 		outDatedDisks = shuffleDisks(outDatedDisks, latestMeta.Erasure.Distribution)
 		partsMetadata = shufflePartsMetadata(partsMetadata, latestMeta.Erasure.Distribution)
 		copyPartsMetadata = shufflePartsMetadata(copyPartsMetadata, latestMeta.Erasure.Distribution)
@@ -823,7 +840,7 @@ func isObjectDangling(metaArr []FileInfo, errs []error, dataErrs []error) (valid
 		break
 	}
 
-	if validMeta.Deleted || validMeta.TransitionStatus == lifecycle.TransitionComplete {
+	if validMeta.Deleted || validMeta.IsRemote() {
 		// notFoundParts is ignored since a
 		// - delete marker does not have any parts
 		// - transition status of complete has no parts

cmd/erasure-object.go

@@ -285,6 +285,11 @@ func (er erasureObjects) getObjectWithFileInfo(ctx context.Context, bucket, obje
 	}
 	var healOnce sync.Once
 
+	// once we have obtained a common FileInfo i.e latest, we should stick
+	// to single dataDir to read the content to avoid reading from some other
+	// dataDir that has stale FileInfo{} to ensure that we fail appropriately
+	// during reads and expect the same dataDir everywhere.
+	dataDir := fi.DataDir
 	for ; partIndex <= lastPartIndex; partIndex++ {
 		if length == totalBytesRead {
 			break
@@ -313,9 +318,8 @@ func (er erasureObjects) getObjectWithFileInfo(ctx context.Context, bucket, obje
 				continue
 			}
 			checksumInfo := metaArr[index].Erasure.GetChecksumInfo(partNumber)
-			partPath := pathJoin(object, metaArr[index].DataDir, fmt.Sprintf("part.%d", partNumber))
-			data := metaArr[index].Data
-			readers[index] = newBitrotReader(disk, data, bucket, partPath, tillOffset,
+			partPath := pathJoin(object, dataDir, fmt.Sprintf("part.%d", partNumber))
+			readers[index] = newBitrotReader(disk, metaArr[index].Data, bucket, partPath, tillOffset,
 				checksumInfo.Algorithm, checksumInfo.Hash, erasure.ShardSize())
 
 			// Prefer local disks
@@ -336,10 +340,12 @@ func (er erasureObjects) getObjectWithFileInfo(ctx context.Context, bucket, obje
 				var scan madmin.HealScanMode
 				if errors.Is(err, errFileNotFound) {
 					scan = madmin.HealNormalScan
+					logger.Info("Healing required, attempting to heal missing shards for %s", pathJoin(bucket, object, fi.VersionID))
 				} else if errors.Is(err, errFileCorrupt) {
 					scan = madmin.HealDeepScan
+					logger.Info("Healing required, attempting to heal bitrot for %s", pathJoin(bucket, object, fi.VersionID))
 				}
-				if scan != madmin.HealUnknownScan {
+				if scan == madmin.HealNormalScan || scan == madmin.HealDeepScan {
 					healOnce.Do(func() {
 						if _, healing := er.getOnlineDisksWithHealing(); !healing {
 							go healObject(bucket, object, fi.VersionID, scan)

cmd/erasure-object_test.go

@@ -324,8 +324,8 @@ func TestGetObjectNoQuorum(t *testing.T) {
 	}
 
 	err = xl.GetObject(ctx, bucket, object, 0, int64(len(buf)), ioutil.Discard, "", opts)
-	if err != toObjectErr(errFileNotFound, bucket, object) {
-		t.Errorf("Expected GetObject to fail with %v, but failed with %v", toObjectErr(errErasureWriteQuorum, bucket, object), err)
+	if err != toObjectErr(errErasureReadQuorum, bucket, object) {
+		t.Errorf("Expected GetObject to fail with %v, but failed with %v", toObjectErr(errErasureReadQuorum, bucket, object), err)
 	}
 
 	// Test use case 2: Make 9 disks offline, which leaves less than quorum number of disks